Unravelling Weather Prediction through Deep Learning and GNSS Approaches in Malawi

The estimation of water vapour in the troposphere holds significant importance for accurate weather forecasting. However, this task is challenged by a limited number of Continuously Operating Reference Station (CORS) stations dedicated to geodynamics in countries like Malawi. In particular, the application of Global Navigation Satellite System (GNSS)-based approaches in disaster-affected areas remains constrained due to the scarcity of CORS monuments. In this presentation, a combined deep learning and the GNSS technique leveraging datasets from the existing CORS network is presented for discussion to exploit the potential of the hybrid arrangement in weather prediction. The proposed approach aims to establish a robust framework for real-time or near real-time weather prediction by considering different weather conditions in Malawi. The talk also highlights the validation results from experimental tests that illustrate the potential applications of such an integrated system.

Keywords: Weather prediction; Deep learning; GNSS; CORS; Malawi